EfficientNet

ConvNet

class lucid.models.EfficientNet(config: EfficientNetConfig)

The EfficientNet class implements a scalable and efficient convolutional neural network architecture that can be configured to encompass all EfficientNet-B0 to B7 variants. Model structure is defined through EfficientNetConfig.

        %%{init: {"flowchart":{"curve":"monotoneX","nodeSpacing":50,"rankSpacing":50}} }%%
flowchart LR
  linkStyle default stroke-width:2.0px
  subgraph sg_m0["<span style='font-size:20px;font-weight:700'>efficientnet_b0</span>"]
  style sg_m0 fill:#000000,fill-opacity:0.05,stroke:#000000,stroke-opacity:0.75,stroke-width:1px
    m1["Upsample"];
    subgraph sg_m2["stage1"]
    style sg_m2 fill:#000000,fill-opacity:0.05,stroke:#000000,stroke-opacity:0.75,stroke-width:1px
      m3["Conv2d<br/><span style='font-size:11px;color:#c53030;font-weight:400'>(1,3,224,224) → (1,32,112,112)</span>"];
      m4["BatchNorm2d"];
    end
    subgraph sg_m5["stage2 x 7"]
    style sg_m5 fill:#000000,fill-opacity:0.05,stroke:#000000,stroke-opacity:0.75,stroke-width:1px
      m5_in(["Input"]);
      m5_out(["Output"]);
  style m5_in fill:#e2e8f0,stroke:#64748b,stroke-width:1px;
  style m5_out fill:#e2e8f0,stroke:#64748b,stroke-width:1px;
      m6["_MBConv<br/><span style='font-size:11px;font-weight:400'>(1,32,112,112) → (1,16,112,112)</span>"];
    end
    subgraph sg_m7["stage9"]
    style sg_m7 fill:#000000,fill-opacity:0.05,stroke:#000000,stroke-opacity:0.75,stroke-width:1px
      m8["Conv2d<br/><span style='font-size:11px;color:#c53030;font-weight:400'>(1,320,7,7) → (1,1280,7,7)</span>"];
      m9["BatchNorm2d"];
      m10["Swish"];
    end
    m11["AdaptiveAvgPool2d<br/><span style='font-size:11px;color:#b7791f;font-weight:400'>(1,1280,7,7) → (1,1280,1,1)</span>"];
    m12["Dropout"];
    m13["Linear<br/><span style='font-size:11px;color:#2b6cb0;font-weight:400'>(1,1280) → (1,1000)</span>"];
  end
  input["Input<br/><span style='font-size:11px;color:#a67c00;font-weight:400'>(1,3,224,224)</span>"];
  output["Output<br/><span style='font-size:11px;color:#a67c00;font-weight:400'>(1,1000)</span>"];
  style input fill:#fff3cd,stroke:#a67c00,stroke-width:1px;
  style output fill:#fff3cd,stroke:#a67c00,stroke-width:1px;
  style m1 fill:#fdf2f8,stroke:#b83280,stroke-width:1px;
  style m3 fill:#ffe8e8,stroke:#c53030,stroke-width:1px;
  style m4 fill:#e6fffa,stroke:#2c7a7b,stroke-width:1px;
  style m8 fill:#ffe8e8,stroke:#c53030,stroke-width:1px;
  style m9 fill:#e6fffa,stroke:#2c7a7b,stroke-width:1px;
  style m10 fill:#faf5ff,stroke:#6b46c1,stroke-width:1px;
  style m11 fill:#fefcbf,stroke:#b7791f,stroke-width:1px;
  style m12 fill:#edf2f7,stroke:#4a5568,stroke-width:1px;
  style m13 fill:#ebf8ff,stroke:#2b6cb0,stroke-width:1px;
  input --> m1;
  m1 --> m3;
  m10 --> m11;
  m11 --> m12;
  m12 --> m13;
  m13 --> output;
  m3 --> m4;
  m4 -.-> m6;
  m5_in -.-> m6;
  m5_out -.-> m5_in;
  m5_out --> m8;
  m6 -.-> m5_in;
  m6 --> m5_out;
  m8 --> m9;
  m9 --> m10;
    

Class Signature

class EfficientNet(nn.Module):
    def __init__(self, config: EfficientNetConfig) -> None

Parameters

• config (EfficientNetConfig): Configuration object describing the compound scaling coefficients, input resolution scale factor, classifier size, squeeze-and-excitation reduction, and optional stochastic depth settings.

Configurations

The following table summarizes the configurations for EfficientNet variants B0 to B7:

EfficientNet Configurations

Variant	Width Coefficient	Depth Coefficient	Input Resolution	Dropout Rate
B0	1.0	1.0	224x224	0.2
B1	1.0	1.1	240x240	0.2
B2	1.1	1.2	260x260	0.3
B3	1.2	1.4	300x300	0.3
B4	1.4	1.8	380x380	0.4
B5	1.6	2.2	456x456	0.4
B6	1.8	2.6	528x528	0.5
B7	2.0	3.1	600x600	0.5

Examples

import lucid
from lucid.models import EfficientNet, EfficientNetConfig

config_b0 = EfficientNetConfig(
    num_classes=1000,
    width_coef=1.0,
    depth_coef=1.0,
    scale=1.0,
)
model_b0 = EfficientNet(config_b0)

input_tensor = lucid.random.randn(1, 3, 224, 224)
output = model_b0(input_tensor)
print(output.shape)

config_b7 = EfficientNetConfig(
    num_classes=1000,
    width_coef=2.0,
    depth_coef=3.1,
    scale=600 / 224,
    dropout=0.5,
)
model_b7 = EfficientNet(config_b7)

input_tensor = lucid.random.randn(1, 3, 600, 600)
output = model_b7(input_tensor)
print(output.shape)